opengear-project / GEAR
GEAR: An Efficient KV Cache Compression Recipefor Near-Lossless Generative Inference of LLM
☆161Updated 9 months ago
Alternatives and similar repositories for GEAR:
Users that are interested in GEAR are comparing it to the libraries listed below
- ☆126Updated 2 months ago
- 16-fold memory access reduction with nearly no loss☆91Updated last month
- [ICML 2024] KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache☆293Updated 3 months ago
- [ICLR 2025] Palu: Compressing KV-Cache with Low-Rank Projection☆103Updated 2 months ago
- [NeurIPS 2024] The official implementation of "Kangaroo: Lossless Self-Speculative Decoding for Accelerating LLMs via Double Early Exitin…☆54Updated 10 months ago
- ☆238Updated last year
- [ICML 2024] Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference☆274Updated 5 months ago
- Boosting 4-bit inference kernels with 2:4 Sparsity☆73Updated 8 months ago
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Serving☆307Updated 10 months ago
- [NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization☆346Updated 8 months ago
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆115Updated 5 months ago
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference☆100Updated 2 weeks ago
- A sparse attention kernel supporting mix sparse patterns☆200Updated 2 months ago
- QAQ: Quality Adaptive Quantization for LLM KV Cache☆49Updated last year
- The Official Implementation of Ada-KV: Optimizing KV Cache Eviction by Adaptive Budget Allocation for Efficient LLM Inference☆72Updated 3 months ago
- Ouroboros: Speculative Decoding with Large Model Enhanced Drafting (EMNLP 2024 main)☆103Updated last month
- PyTorch implementation of paper "Response Length Perception and Sequence Scheduling: An LLM-Empowered LLM Inference Pipeline".☆85Updated last year
- [ICLR 2025] PEARL: Parallel Speculative Decoding with Adaptive Draft Length☆80Updated 3 weeks ago
- ☆53Updated last year
- Code associated with the paper **Draft & Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding**☆184Updated 2 months ago
- The official implementation of the paper "Towards Efficient Mixture of Experts: A Holistic Study of Compression Techniques (TMLR)".☆66Updated last month
- PyTorch bindings for CUTLASS grouped GEMM.☆87Updated this week
- The official implementation of the EMNLP 2023 paper LLM-FP4☆198Updated last year
- The official implementation of the paper <MoA: Mixture of Sparse Attention for Automatic Large Language Model Compression>☆127Updated 5 months ago
- [ICML 2024 Oral] Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs☆106Updated 2 weeks ago
- An algorithm for weight-activation quantization (W4A4, W4A8) of LLMs, supporting both static and dynamic quantization☆129Updated 3 months ago
- ☆73Updated 2 weeks ago
- Official Pytorch Implementation of "Outlier Weighed Layerwise Sparsity (OWL): A Missing Secret Sauce for Pruning LLMs to High Sparsity"☆65Updated 10 months ago
- Activation-aware Singular Value Decomposition for Compressing Large Language Models☆66Updated 6 months ago
- [ICLR 2025] COAT: Compressing Optimizer States and Activation for Memory-Efficient FP8 Training☆189Updated 2 weeks ago